Systems and methods for deriving quantization parameters for video blocks in video coding

Abstract

Some embodiments provide a method of decoding a current video block in a video picture. From several partitioning schemes, the method identifies a partitioning scheme that is used to partition the current video block. Based on the identified partitioning scheme, the method selects one derivation to compute a first quantization parameter for the current video block from several candidate derivations for computing the first quantization parameter. The method computes a second quantization parameter for the current video block based at least in part on the computed first quantization parameter. The method then uses the computed second quantization parameter to decode the current video block.

The invention claimed is: 
     
       1. A method of decoding a current video block in a video picture, the method comprising:
 determining whether a partitioning scheme that is used to partition the current video block is a single tree partitioning that uses one partitioning for both luma and chroma color components of the current video block or a dual tree partitioning that uses two different partitionings for the luma color component of the current video block and the chroma color component of the current video block; 
 based on whether the determined partitioning scheme is single tree partitioning or dual tree partitioning, selecting one derivation to compute a first predictive quantization parameter for the current video block from a plurality of candidate derivations for computing the first predictive quantization parameter, wherein (i) a first derivation for computing the first predictive quantization parameter when the partitioning scheme is single tree partitioning comprises computing the first predictive quantization parameter based at least in part on another quantization parameter and an additional value and (ii) a second derivation for computing the first predictive quantization parameter when the partitioning scheme is dual tree partitioning comprises computing the first predictive quantization parameter by using a previously-derived quantization parameter; 
 computing a second quantization parameter for the current video block based at least in part on the computed first predictive quantization parameter and an offset value; and 
 using the computed second quantization parameter to decode the current video block. 
 
     
     
       2. The method of  claim 1  further comprising performing a quantization operation on values associated with the current video block using the computed second quantization parameter. 
     
     
       3. The method of  claim 1 , wherein using the computed second quantization parameter comprises performing an inverse quantization operation on values associated with the current video block using the computed second quantization parameter. 
     
     
       4. The method of  claim 1 , wherein:
 the first derivation is selected when the single tree partitioning scheme is identified as the partitioning scheme for the current video block and for a reference video block that is used for encoding the current video block; and 
 the second derivation is selected when the dual tree partitioning scheme is identified for the current video block and a single tree partitioning scheme is identified for the reference video block. 
 
     
     
       5. The method of  claim 4 , wherein the reference video block is a neighboring coded video block. 
     
     
       6. The method of  claim 5 , wherein the neighboring coded video block is a video block to the left of the current video block in the current picture that includes the neighboring and current video blocks. 
     
     
       7. The method of  claim 5 , wherein the neighboring coded video block is a video block above the current video block in the current picture that includes the neighboring and current video blocks. 
     
     
       8. The method of  claim 4 , wherein the reference video block is a last coded video block in the current picture that includes the neighboring and current video blocks. 
     
     
       9. The method of  claim 1 , wherein the first derivation uses at least one equation not used by the second derivation. 
     
     
       10. The method of  claim 1 , wherein the computed first and second quantization parameters are first and second quantization parameters for the luma color component of the current video block. 
     
     
       11. The method of  claim 1 , wherein the computed first and second quantization parameters are first and second quantization parameters for the chroma color component of the current video block.